I am new person
You joined the site over a year ago. That's not new.
Chess IS solved. Chess is a finite game, so it is POSSIBLE to solve.
That part is correct. Finite - so solvable.
But that doesn't mean it will ever be solved.
We do see a claim repeated over and over -
that its 'solvable' because of a claim within claim as to a time it would be done in of less than five years.
That's analagous to claiming that Mars could be developed because its just a matter of putting the money into it so that spaceships could do it in ten years?
No - its worse than that. Because spaceships developing the planet Mars looks less 'unrealistic' while still being intensely unrealistic.
Computers just aren't fast enough yet. To 'solve' chess.
If they're 1000 times faster ten years from now - that doesn't shave enough zeroes off the time needed. Would still take gazillions of years.
There's all kinds of manoeuvering though.
Like 'taking the square root of the number of possible positions'
😁😁😁
But these conversations survive in the form of 'weakly analyzed' or 'weakly solved'. Whatever.
Chess IS solved. Chess is a finite game, so it is POSSIBLE to solve. Analyzed by humans, no probably not. By computers, it already has been analyzed.
I agree that it is possible to solve, but I don't think it has been solved yet.
I hope it never is solved though, it will suck if we reach the point where Stockfish ∞ announces white to mate in 63 moves, before the game has started.
#102
"it will suck if we reach the point where Stockfish ∞ announces white to mate in 63 moves"
++ Do not worry, it is: white to play, back draws.
#103
"Maybe quantum computers can solve it"
++ To strongly solve chess needs to evaluate all 10^44 legal positions.
That would require a 146 qubit quantum computer.
There already exist larger quantum computers:
https://www.dwavesys.com/company/newsroom/media-coverage/hpcwire-d-wave-previews-next-gen-platform-debuts-pegasus-topology-targets-5000-qubits/
Weakly solving chess is already within reach of conventional computers:
it takes 5 years on 3 cloud engines of 10^9 nodes/s.
The number of legal, sensible, reachable, and relevant positions is only a tiny fraction of the number of legal positions.
"Weakly solving chess is already within reach of conventional computers:
it takes 5 years on 3 cloud engines of 10^9 nodes/s."
It takes5 seconds - one can simply say 'weakly' that chess games end in a draw unless somebody makes a mistake.
But its 'weakly' because that hasn't been proven.
"The number of legal, sensible, reachable, and relevant positions is only a tiny fraction of the number of legal positions."
But that fraction isn't known. So 'five years' is invalid.
And 'nodes' in "10^9 nodes" isn't defined as to what that means as far as solving speed is concerned - so that's invalid too.
And even a 'tiny fraction' like one in ten thousand isn't going to rescue the computers from a number of possible positions that is far too large.
What would take 'five years' is a phony project dressed up to look like its solving something but isn't.
Which is a probable reason why nobody has put up the money for it.
Who would pay for phony 'solving'?
Finally - there's the issue of spreading phony information around on the internet to the effect 'it takes five years'.
Well its a bit like claiming the earth is flat.
Which is harmless because people know its wrong - or can harmlessly be duped into that. Its like UFO belief.
A trip on a boat or ship along a coast will demonstrate the receding horizon. But even that kind of proof isn't necessary because people using commonsense know that a round earth couldn't be faked.
None of these look dangerous though - unlike Covid disinformation and misinformation of various kinds. Those are harmful.
Chess IS solved. Chess is a finite game, so it is POSSIBLE to solve. Analyzed by humans, no probably not. By computers, it already has been analyzed.
I agree that it is possible to solve, but I don't think it has been solved yet.
I hope it never is solved though, it will suck if we reach the point where Stockfish ∞ announces white to mate in 63 moves, before the game has started.
Maybe the game doesn't need to have every single possible position analyzed, if a sequence of moves can be found that will force checkmate right from the start, no matter what is played to try and stop it...
#115
"the overwhelming number of possible games of chess are nonsense in chess terms"
++ Yes, that is right. Besides it is not every possible sequence, but every legal, sensible, reachable, and relevant position.
Many sequences transpose to the same position: 1 e4 e5 2 Bc4 Nc6 3 Nf3 = 1 e4 Nc6 2 Nf3 e5 3 Bc4 = 1 Nc3 Nf6 2 e4 e5 3 Nf3 Nc6 4 Bc4 Ng8 5 Nb1 etc. etc.
Many positions are nonsense, like with a material deficit without compensation.
#117
"in the process of trying to analyse chess, it's necessary to prune search trees"
++ Yes, very much so.
It is necessary to prune search trees.
It is necessary to end search when a known draw or win has occured even with > 7 men.
"it is often beneficial to incorporate knowledge-based methods in game-solving programs"
- Prof. Em. J. van den Herik
"it is often beneficial to incorporate knowledge-based methods in game-solving programs"
which is confused with unfounded skipping of positions.
Or used instead of confused - to justify such illogical skipping.
And mentioning a name like den Herik doesn't make something right either.
And apparently nobody is buying - in the literal sense of the word that there's a legitimate 'weakly' solving of chess in five years.
ches is a very comlicated game
computers even though they are really good
chess is way too many numbers for themm
ches is a very comlicated game
computers even though they are really good
chess is way too many numbers for themm
Current computers are very fast - but just not nearly fast enough to 'solve' chess.
If they were a trillion times faster - then maybe a chance?
But how long would that take? For computers to improve that much ?
That in itself could take thousands of years - or even millions.
ches is a very comlicated game
computers even though they are really good
chess is way too many numbers for themm
Current computers are very fast - but just not nearly fast enough to 'solve' chess.
If they were a trillion times faster - then maybe a chance?
But how long would that take? For computers to improve that much ?
That in itself could take thousands of years - or even millions.
Possible...
Though chess is a very infinte game...
It will take a lot of strong programmers to take on such a feat...
My age is 7